Learning is a level-progressing process. In any field of study, one must master basic concepts to understand more complex ones. Thus, it is important that during the learning process learners are presented and challenged with knowledge which they are able to comprehend (not a level below, not a level too high). In this work we focus on language learners. By gradually improving (complicating) texts, readers are challenged to learn new vocabulary. To achieve such goals, in this paper we propose and evaluate the 'complicator' that translates given sentences to a chosen level of higher degree of difficulty. The 'complicator' is based on natural language processing and information retrieval approaches that perform lexical replacements. 30 native English speakers participated in a user study evaluating our methods on an expert-tailored dataset of children books. Results show that our tool can be of great utility for language learners who are willing to improve their vocabulary. The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-642-40814-4_45.